PROJECT SUMMARY/ABSTRACT Prostate cancer (PC) is the second most common cancer in American men and second most lethal. Options for treating metastatic castration-resistant PC (mCRPC) are limited and new therapeutic strategies are needed. About 20% of mCRPC patients have tumors harboring mutations in DNA repair genes, mostly in the homologous recombination (HR) pathway (e.g. BRCA2). About 5% of mCRPCs have mutations in CDK12, a cyclin-dependent kinase whose loss leads to aberrant splicing and mRNA downregulation of many DNA repair genes. These DNA repair-deficient tumors tend to be very aggressive but can be treatable, as loss of HR genes (e.g. BRCA2) is known to sensitize other cancers (e.g. ovarian) to poly(ADP) ribose polymerase inhibitors (PARPi) and platinum chemotherapy (PLAT). However, it is unknown if CDK12 loss sensitizes mCRPC to this therapy. Furthermore, CDK12-mutant mCRPC show recurrent amplifications in cell cycle genes (e.g. cyclin D1) and rarely carry other HR mutations or upregulation of ETS-family genes (e.g. TMPRSS2-ERG fusions). This proposal will develop new in vitro and in vivo models to investigate the precise role of DNA repair mutations in PC development by testing three hypotheses/specific aims: (1) determine if loss of Brca1, Brca2, or Cdk12 is sufficient to drive or accelerate prostate tumorigenesis and identify cooperating genomic aberrations that contribute to a penetrant phenotype, (2) determine if CDK12-mutant PCs respond to therapies targeting DNA repair and cell cycle checkpoints, and (3) determine the efficacy of targeting CDK12 in HR-deficient and ETS+ prostate cancers. I will generate conditional genetic mouse models with prostate-specific deletion of Brca1, Brca2, or Cdk12 to test if their loss is sufficient for PC development. Mice will also be crossed to Ptenfl/fl or Pb-MYC mice to test if DNA repair loss can accelerate tumor development in established mouse models of PC. Tumors will be analyzed by DNA and RNA sequencing to determine if they recapitulate genomic alteration patterns as seen in human tumors. This project will also utilize PC cell lines engineered with inducible BRCA2 and CDK12 knockdown and patient derived xenografts (PDX) with endogenous mutations to test if loss of CDK12 sensitizes PC to PARPi+PLAT (olaparib+cisplatin) or CDK4/6 inhibitor (ribociclib). Lastly, engineered PC lines or PDX lines with endogenous BRCA2 loss or TMPRSS2-ERG fusion will be used to test if CDK12 knockdown or inhibition (THZ531) is synthetically lethal in cells with HR loss or ETS-family oncogene upregulation. Together, these studies will address key questions concerning the role of DNA repair loss in PC and test potential targeted therapies for CDK12-mutant PC. Furthermore, these experiments will determine if targeting CDK12 can be effective for key subsets of PC, which could spur future trials and development of CDK12 inhibitors and potentially offer new therapeutic options to extend survival for many mCRPC patients.